Automatic Validation Principles
Reference | Automatic Validation
Minimising boilerplate code
Implementing validation of user data submitted to web services generally results in lot of boilerplate code, especially in a statically typed language like Go that doesn’t support concepts like annotations.
A primary goal of Granitic’s validation framework is to remove as much of this boilerplate code as possible.
Validation as rules
Rather than implementing validation as imperative code, Granitic allows the various checks (operations, in Granitic terminology) performed on a field submitted by a user to be expressed as a terse rule.
These rules are defined in Granitic’s standard configuration files and as such can be modified without recompiling your application and can have their definition changed between envrionments, if required.
Integration with processing pipeline
Automatic validation is fully integrated with the web service processing pipeline. Its goal is to deliver a validated data capture target to your application logic or automatically return an error response if validation failed.
Custom operations
Granitic includes a number of common operations for checking various data-types, but also provides a mechanism for you to build custom operations for additional checks (e.g. making sure an email address is not already being used by another user)